From ad7a8ad8793019ad93cee971187cc7d0f0cca462 Mon Sep 17 00:00:00 2001
From: Raptorox <70806316+Raptorox@users.noreply.github.com>
Date: Wed, 20 May 2026 21:04:14 +0200
Subject: [PATCH] add lexer

---
 .gitignore   |  1 +
 src/lexer.rs | 96 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 src/main.rs  | 25 ++++++++++++--
 3 files changed, 120 insertions(+), 2 deletions(-)
 create mode 100644 src/lexer.rs

diff --git a/.gitignore b/.gitignore
index ea8c4bf..767ad53 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1 +1,2 @@
 /target
+/res
\ No newline at end of file
diff --git a/src/lexer.rs b/src/lexer.rs
new file mode 100644
index 0000000..d49eafd
--- /dev/null
+++ b/src/lexer.rs
@@ -0,0 +1,96 @@
+#[derive(Debug, PartialEq)]
+pub enum Token {
+    Number(i64),
+    Ident(String),
+
+    // Operators
+    Add,
+    Sub,
+    Mul,
+    Div,
+    Mod,
+
+    // Parentheses
+    LParen,
+    RParen,
+    LBrace,
+    RBrace,
+    LBracket,
+    RBracket,
+
+    Semicolon
+}
+
+pub struct Lexer {
+    source: Vec<u8>,
+    pos: usize
+}
+
+impl Lexer {
+    pub fn new(source: &str) -> Self {
+        Self {
+            source: source.as_bytes().to_vec(),
+            pos: 0
+        }
+    }
+
+    fn skip_whitespace(&mut self) {
+        while matches!(self.peek(), Some(c) if c.is_ascii_whitespace()) {
+            self.advance();
+        }
+    }
+
+    fn advance(&mut self) {
+        self.pos += 1;
+    }
+
+    fn peek(&self) -> Option<u8> {
+        self.source.get(self.pos).copied()
+    }
+
+    fn number(&mut self) -> Option<Token> {
+        let mut num = String::new();
+
+        while matches!(self.peek(), Some(b'0' ..= b'9')) {
+            num.push(self.peek().unwrap() as char);
+            self.advance();
+        }
+
+        Some(Token::Number(num.parse().unwrap()))
+    }
+
+    fn ident(&mut self) -> Option<Token> {
+        let mut ident = String::new();
+
+        while matches!(self.peek(), Some(b'a' ..= b'z' | b'A' ..= b'Z')) {
+            ident.push(self.peek().unwrap() as char);
+        }
+
+        Some(Token::Ident(ident))
+    }
+
+    pub fn next(&mut self) -> Option<Token> {
+        self.skip_whitespace();
+        match self.peek() {
+            Some(b'0' ..= b'9') => self.number(),
+            Some(b'a' ..= b'z' | b'A' ..= b'Z') => self.ident(),
+
+            Some(b'+') => {self.advance(); Some(Token::Add)},
+            Some(b'-') => {self.advance(); Some(Token::Sub)},
+            Some(b'*') => {self.advance(); Some(Token::Mul)},
+            Some(b'/') => {self.advance(); Some(Token::Div)},
+            Some(b'%') => {self.advance(); Some(Token::Mod)},
+
+            Some(b'(') => {self.advance(); Some(Token::LParen)},
+            Some(b')') => {self.advance(); Some(Token::RParen)},
+            Some(b'[') => {self.advance(); Some(Token::LBracket)},
+            Some(b']') => {self.advance(); Some(Token::RBracket)},
+            Some(b'{') => {self.advance(); Some(Token::LBrace)},
+            Some(b'}') => {self.advance(); Some(Token::RBrace)},
+
+            Some(b';') => {self.advance(); Some(Token::Semicolon)}
+            
+            _ => None
+        }
+    }
+}
\ No newline at end of file
diff --git a/src/main.rs b/src/main.rs
index e7a11a9..59ed221 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -1,3 +1,24 @@
-fn main() {
-    println!("Hello, world!");
+use std::{fs::File, io::Read};
+
+mod lexer;
+use lexer::Lexer;
+
+fn main() -> std::io::Result<()> {
+    let args = std::env::args().collect::<Vec<String>>();
+
+    let mut source = String::new();
+    File::open(&args[1])?.read_to_string(&mut source)?;
+
+    println!("{:?}", source);
+
+    let mut lexer = Lexer::new(&source);
+
+    loop {
+        match lexer.next() {
+            Some(tok) => print!("{tok:?}, "),
+            None => break
+        }
+    }
+
+    Ok(())
 }